BUG/MEDIUM: mworker: fix FD leak upon reload

We reintroduced some FDs leaking by using a poller and some listeners in
the master.

The master proxy needs to be stopped to avoid leaking its listeners, the
polling loop needs to be deinit, and the thread waker pipe need to be
closed too.

No backport needed.
This commit is contained in:
William Lallemand 2018-11-26 11:53:40 +01:00 committed by Willy Tarreau
parent e548974ca8
commit 7c756a8ccc
1 changed files with 13 additions and 0 deletions

View File

@ -704,6 +704,7 @@ static void mworker_reload()
int next_argc = 0;
int j;
char *msg = NULL;
struct per_thread_deinit_fct *ptdf;
mworker_block_signals();
#if defined(USE_SYSTEMD)
@ -714,6 +715,18 @@ static void mworker_reload()
mworker_proc_list_to_env(); /* put the children description in the env */
/* during the reload we must ensure that every FDs that can't be
* reuse (ie those that are not referenced in the proc_list)
* are closed or they will leak. */
/* close the listeners FD */
mworker_cli_proxy_stop();
/* close the poller FD and the thread waker pipe FD */
list_for_each_entry(ptdf, &per_thread_deinit_list, list)
ptdf->fct();
if (fdtab)
deinit_pollers();
/* compute length */
while (next_argv[next_argc])
next_argc++;